By-election triple bill

Something for everyone?

Last week was an exciting one for election nerds like your Swingometer host, with three by-elections in Conservative held English seats all on the same night. The results delivered one seat each for the three main parties - the Lib Dems took the Somerset seat of Somerton and Frome on a whopping 29 point swing, Labour took the Yorkshire seat of Selby and Ainsty on a massive 24 point swing, and the Conservatives held on against the tide in the London marginal of Uxbridge and South Ruislip, with Labour only able to muster a disappointing 6.7 point swing.

Something for everyone to cheer about, then, but while the surprise Uxbridge win will boost Downing Street morale, with Rishi Sunak averting the grim fate of becoming the first Prime Minister since Harold Wilson in 1968 to lose three by-elections on the same day¹, it is the opposition parties who can be happiest with the trio of results.

Results from the 20th July by-election triple bill

There is clear evidence of tactical voting at work in these contests - the Lib Dem vote plunged in both the seats Labour was challenging, while the Labour vote collapsed in Somerton. This adds to a growing body of evidence that Labour and Lib Dem voters are co-ordinating behind whoever is best placed to oust local Conservatives, an ominous trend for Team Sunak.

Despite a lot of sound and fury about the electoral potency of an “anti-woke” agenda in the political media, there not much evidence of voter enthusiasm for the keenest proponents of ‘culture war’ politics at these by-elections. The fringe right ReformUK and Reclaim parties both flopped once again in all three contests, continuing a long run of weak showings for the radical right.

The turnout decline was much smaller in Uxbridge than the other two contests, which is consistent with the theory that annoyance at ULEZ may have motivated Con voters otherwise unhappy with the govt or apathetic about their party to turn out and register their disapproval of the local Labour mayor and his policies.

On the other hand, the Greens were up in all three contests, and up sharply in Somerton. Watering down Green commitments, an idea which has gained traction among commentators interpreting Uxbridge as a voter revolt against costly green policies, may not be a risk free prospect. While some voters may have cast protest ballots againts ULEZ, others are clearly eager to register their enthusiasm for the much more radical agenda of the Green party.

Sunrise in the South West for the Lib Dems

Victory over the Conservatives on massive swings has become routine for the Liberal Democrats in by-elections over the past couple of years. Somerton and Frome was the fourth Tory gain on swings of 25 points or more in the past two years. It is also the second by-election gain this Parliament in the South West, which before Coalition was a heartland for the third party. The South West returned 15 Lib Dem MPs in 2010, more than a quarter of the party’s total. Every single one lost their seat in the post-Coalition bloodbath of 2015. The Lib Dems’ strong Remain stances delayed their post-Coalition recovery and by 2019 the Lib Dems still had only one seat in the region (Bath). Yet, as the table below reveals, old loyalties have endured in a region which leant towards the Liberals for generations, and the Lib Dems remain in second place to the Tories in over 20 South West seats.

There are eleven Conservative seats in the South West region with majorities over the Lib Dems which are smaller than Somerton and Frome, and another eight which would fall on swings of 21 points or less, well below the 29 point swing achieved last week. Fourteen of these seats were held by the Lib Dems pre-coalition, including the Yeovil seat once represented by long serving party leader Paddy Ashdown. Tactical voting could also provide a substantial boost, with big third placed Labour votes available to squeeze in many seats.

While whopping swings can’t be expected everywhere, these are clearly areas with a historical and cultural affinity for the Lib Dems, raising the chances of some big rebounds. Local Lib Dem strength is already returning, with major gains in South West local elections this year and last returning the third party to power in many local councils. With memories of Coalition fading, and Brexit no longer a barrier, many of these seemingly deep blue seats could be vulnerable to a Lib Dem restoration next year.

Conservative-Lib Dem battlegrounds in the South West region

Back to the 1990s for Labour

Labour’s performance in Selby and Ainsty was exceptionally good. This was the second largest swing to Labour ever recorded in a by-election, and one of only half a dozen swings from Conservative to Labour of over 20 points ever recorded. Four of the other five were achieved by Tony Blair between 1994 and 1996 en route to a landslide general election. The fifth came in Mid-Staffordshire in the spring of 1990, when Neil Kinnock opposed a deeply unpopular Margaret Thatcher government, in what proved to be Labour’s last by-election gain from the Conservatives of Mrs Thatcher’s long premiership.²

Swings of over 20 points to Labour in by-elections since 1945

Team Starmer can therefore point to Selby as evidence that Labour are returning to peak Blair levels of electoral strength less than four years after being reduced to their worst Commons showing since 1935. Yet, much like Blair in opposition, Starmer and his allies seems more focussed on the disappointing result in Uxbridge than the dominant one in Selby. The Uxbridge swing, at 6.7 points, is indeed well below that seen in current polling and well below what Labour would need to secure a majority at the next general election. But there are at good reasons to think Uxbridge may be an outlier for Labour.

Uxbridge has a long history of resisting Labour advances, suggesting an unusually solid Conservative vote. Hillingdon, its London borough, has continued to return Conservative council majorities even as the capital as a whole has swung strongly towards Labour, and the area swung against Labour in the 2021 London Mayoral and Assembly elections. Labour has failed twice before in seemingly winnable by-elections in Uxbridge, a patch of London which though often marginal has only returned a Labour MP once since 1959. The voters of Uxbridge stuck with the Conservatives in 1972, on the same day as London voters a little further south in Sutton and Cheam elected a Liberal on a 33 point swing. In 1997, Uxbridge resisted the advances of New Labour just a few months after Tony Blair’s first landslide, with a swing of 5 points to the Conservatives.

There was also an unusual dynamic to this election, with many voters seeing the by-election not as the typical opportunity to register discontent with the Westminster government, but instead a chance to voice a protest at the policies of the Labour administration in City Hall. London’s Ultra Low Emission Zone is slated to expand next month, taking in outer borough areas including Uxbridge. A lot of motorists were vocally unhappy about it, something both parties picked up on the campaign trail. As a consequence, the result is now being seen as a case study in the electoral risks of environmental policies.

While some sort of ULEZ effect is plausible, it is not the only potential explanation for Labour under-peformance in Uxbridge. This is also a seat with a substantial Indian origin population (13% of residents in the 2021 census) and there have been several pieces of evidence recently that areas with large Indian populations may be shifting towards the Conservatives. There was a Labour to Tory swing in the London borough of Harrow in 2022, the only borough picked up by the Conservatives while most of London swung to Labour. Harrow has the largest Indian origin population of any London borough. Something similar happened in local elections this year: Leicester, the city with the largest Indian origin population outside London, swung heavily to the Conservatives, again running against the national pro-Labour current. And Leicester East, which has the country’s highest share of Indian origin voters, had by far the largest anti-Labour swing in an ethnically diverse seat in the 2019 general election.³

There are a number of plausible reasons why Indian voters might be shifting in a Tory direction even as other ethnic minority groups remain staunchly Labour and white voters swing against the government. Rishi Sunak is the first Prime Minister of Indian origin, something which may resonate strongly with voters sharing a similar background. The British Indian community is also relatively prosperous and middle class, factors which may erode Labour’s historic advantages. Perceptions that Labour is critical of, and the Conservatives friendly towards, the current Modi government in India may be another factor, as may reforms to immigration rules which have made it easier for highly qualified Indians in professional jobs to settle in England. All of this, however, is somewhat speculative as at present we have very little reliable polling data available to examine shifts in Britain’s ethnic minority communities.

However, even if a blue shift in the Indian community is underway, this has limited broader upside for the Conservatives in the next general election as most of the seats with large Indian populations are safe Labour seats where a gain in this community alone will not be enough to change the political weather. But keep an eye next year on Harrow East, where Indian heritage votes might help the Conservatives cling on against the tide, and Leicester East, a newly marginal seat with a majority Indian heritage electorate.

Are Labour now on course for victory?

By-elections are not perfect indicators of future election success, but as I’ve discussed in earlier Swingometer posts, they do provide us with some clues, particularly if we focus on “big two” contests between the Conservatives and Labour. The chart below shows average swings from the government to the opposition in every parliament since 1945.⁴ Blue bars signal swings to the Conservatives during Labour governments, red bars (more frequent) signal swings to Labour during Tory terms. Dark bars mean the opposition won the next election, light bars mean the government prevailed. Labour’s average for the whole Parliament as a whole is in grey at the end; the opposition’s average since the “partygate” scandals broke is the final black bar.

Labour’s average swing for the Parliament as a whole is, at 7 points, still below what oppositions typically need. Ed Miliband did just about as well in “big two” contests in 201-15 and still lost, Neil Kinnock did better in 1983-1987 and 1987-1992 before going on to defeat. If, however, we focus on the period since “partygate” triggered a major shift in public opinion against the government, the average swing now stands at 10.5 points. This is very similar to the swings achieved by David Cameron and Margaret Thatcher in the terms before they took their parties into government, though some way behind Tony Blair’s record average of 14.4 points.

Average government to opposition swings, “big two” by-elections, Parliaments since 1945

In the graph below, I make the same point even clearer by taking the overall average for all the postwar terms when the incumbent party goes on to win (light grey) and when the opposition wins (black). By-election swings in terms when the incumbent ends up prevailing average 4 points; when the opposition wins the average swing is nine points. Labour’s overall average for the term - 7 points - is below the average swing winning oppositions have manged. But since partygate, Labour have been well above the past average for oppositions heading to victory.

Average swings to the opposition since 1945 in Parliaments where the incumbent wins (grey) and when the opposition wins (black), and Labour’s performances since 2019 (light red) and since December 2021 (dark red)

More good news to come for the opposition?

The split result in Selby and Uxbridge is, in a way, a fair representation of the overall story for Labour so far so far this term: generally good, and getting better, but still not quite good enough to dispel all doubts. Further demonstrations of strength would help convince the sceptics, and opportunities for these may not be long in coming, with perhaps three more by-elections on the cards before the year is out. Conservative MP Nadine Dorries has announced an intention to resign her Mid-Bedfordshire seat, though she is yet to act on it. Her former colleague Chris Pincher, whose misdemeanours started the storm which finally blew Boris Johnson out of office, may soon be ejected in turn from his Tamworth seat. His colleagues on the standards committee have recommended he be suspended from Parliament for eight weeks, a penalty severe enough to trigger a recall petition which, if successful (and most have been) will result in a by-election. And there is a recall petition already underway in former SNP MP Margaret Ferrier’s seat, following her serious breaches of COVID regulations.⁵

Swing achieved by Labour in Selby and Ainsty, and swing required in three potentially up-coming by-elections

The two Tory seats provide opportunities to prove Selby was no accident - both would fall to Labour on swings smaller than the opposition achieved in Selby. Or Mid-Bedfordshire could provide another whopping Lib Dem swing - the third party have targeted the seat early, and may be hoping the two upcoming contests can be split between parties as last week’s were. Rutherglen and Hamilton West, meanwhile, provides a test of Scottish Labour’s newfound polling strength in what should be an easily winnable SNP marginal seat. However, as we saw last week, the easiest contests on paper don’t always turn out that way in practice.

This post was edited at 12pm on 25th July to correct an error regarding Uxbridge - contrary to my initial claim it has elected Labour MPs in the past, though Labour has prevailed only once in general elections or by-elections since 1959 (a narrow win in 1966). My thanks to Graham Harries for flagging the error.

1. Wilson lost Acton, Dudley and Meriden on 28th March 1968, all on very large swings. One of those first elected that day was the long serving Conservative MP, minister and peer Kenneth Baker. Dudley would go on to set another record when Tony Blair achieved the largest ever swing to Labour in the same town in 1994. The largest ever Labour to Conservative swing also happened in a West Midlands seat, Walsall North (1976), while two of the other five largest Conservative to Labour swings happened in nearby Staffordshire.

2. There were five more by-elections in mainland Britain before Mrs Thatcher’s departure at the end of 1990. Four of these were in very safe Labour seats, which put a ceiling on the swings Labour could achieve, including two within 6 months in Bootle, where the candidate elected in May died less than two months later. The fifth by-election was the contest to replace Ian Gow in Eastbourne following his death at the hands of the IRA - the last MP to be killed in service until the murder of Jo Cox in 2016. The Liberal Democrats won a famous victory on a 20 point swing.

3. Leicester East was for thirty years the seat of Keith Vaz, one of the first four BME MPs elected in 1987. Scandal prompted Vaz to step down in 2019, and he was replaced by close Corbyn ally and Islington councillor Claudia Webbe. A fifteen point swing to the Conservatives slashed Labour’s majority from 43% to just 12%.

4. February to October 1974 is omitted - it isn’t long enough to generate a meaningful average

5. The Rutherglen and Hamilton West petition closes on 31st July, so the wait for another by-election may not be a long one.